## 4.1 Identifying Shocks to State Capacity and Institutions

## Source details

**Canonical URL:** [4.1 Identifying Shocks to State Capacity and Institutions](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025117-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025117-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025117-print-pdf.pdf.json)

---

### Identification strategy and shock construction
- Two complementary shock definitions:
  - Residual-based shocks: estimate expected evolution of each indicator with country fixed-effect autoregressive models augmented with log real GDP per capita and log population:
    - log(GOV_it) = α_i + β1 log(GOV_it−1) + β2 log(GDP_it) + β3 log(POP_it) + u_it
    - Extract residuals u_it and classify a discrete shock when u_it > ū_i + σ_i (country-specific mean plus one standard deviation).
  - Percentile-based shocks: define a shock as an annual change in the indicator that exceeds the country-specific 90th percentile of its historical annual-change distribution (top 10 percent).
- Scaling:
  - Local projection coefficients are scaled by the average magnitude of the governance change to express results in terms of a typical large institutional shift.
- Instruments / identification checks:
  - SC shocks instrumented with natural disasters with high fatality rates as exogenous triggers of state-building episodes.
  - IN shocks (GEE and PRCL) instrumented with the average annual change in Polity2 scores among countries within the same IMF income group (excluding the target country), exploiting regional institutional diffusion.

### Rationale and identification challenges
- Motivation:
  - SC and IN indicators are highly persistent and evolve slowly; simple year-on-year changes risk capturing endogenous dynamics.
  - The multi-step procedure isolates large, unexpected deviations more plausibly exogenous to contemporaneous macroeconomic conditions, improving causal inference.
- Empirical strengths:
  - Dual shock strategy captures both large residual-driven deviations and relatively rare but sizable within-country annual changes.
  - Aligns with literature emphasizing macroeconomic relevance of large institutional shifts.

---

### 4.2 Local Projections — Methodology and Data

### Local Projections (LP) specification and implementation
- Core LP regression (notation preserved):
  - ∆log y_{i,t+h} = α_{i,h} + δ_{t,h} + ∑_{j=0}^{L} β_{j,h} d_{i,t−j}^2 + ∑_{l=0}^{L} β_{l,h} (log y_{i,t−l} − log y_{i,t−1−l}) + ∑_{h=1}^{H} β_h d_{i,t+h} + ∑_{c=0}^{C} β_{c,h}′ X_{i,t−c} + u_{i,t+h}
- Key parameterization and facts:
  - Forecast horizon h = 1, ..., 8.
  - IRFs scaled by the average change in the relevant indicator when shocks occur.
  - Maximum of 3,826 observations for state capacity changes for one-year-ahead forecasts; lose up to 129 observations per additional year due to panel structure.
  - For institutional proxies: 1,126 observations corresponding to 55 countries (GEE) and 4,719 observations corresponding to 130 countries (PRCL) in respective specifications.
- Lag selection and stationarity:
  - AIC indicates 2 lags of the treatment variable and 1 lag of ∆log y.
  - Data stationary: |∑_{l=0}^4 β_{l,h}| < 1 in all specifications.
- Overlapping-horizon bias correction:
  - Include leads ∑_{h=1}^H β_h d_{i,t+h} to capture Teulings and Zubanov (2014) correction; leads are statistically significant for most combinations.
- Standard errors and cross-sectional dependence:
  - Use Spatial Correlation Consistent (SCC) / Driscoll-Kraay (1998) standard errors.
  - Pesaran (2015) test often detects cross-sectional dependence; SCC SEs mitigate concerns.

### State-dependent LPs and STAR augmentation
- State-dependent (nonlinear) LP specification augments LPs with a smooth transition weighting:
  - Weighting function F(z_it) = exp(−γ z_it) / (1 + exp(−γ z_it)), γ > 0; z_it is the conditioning variable normalized to zero mean and unit variance.
  - Constructs low- and high-regime IRFs via β^L and β^H weighted by F(z_it) and 1 − F(z_it).
- Advantages:
  - LPs are flexible for panel state-dependence, avoid VAR-to-VMA inversion complications, and do not require assumptions on duration of states.
- Caveat:
  - If the state is correlated with the shock, LPs recover responses only to small shocks; in this sample, unconditional correlation between SC shocks and IN shocks is very low and statistically insignificant at the 5 percent level.

### Data and measurement
- State Capacity (SC): Hanson and Sigman (2021) latent-index integrating 21 indicators (extractive, coercive, administrative); dataset spans 177 countries, 1960–2015.
- Institutional quality (IN):
  - Government Effectiveness Index (GEE), World Bank WGI: over 200 countries since the 1990s.
  - Political Civil Liberties Index (PRCL), V-Dem (processed by Our World in Data): spans 1789–2024 (central estimate).
- Control vector X_{i,t} includes contemporaneous and first lags of:
  - Investment (Gross Capital Formation), Inflation Rate, Trade Openness, Population Growth, Policy Uncertainty.

---

### 6.1 Baseline LP Results — Main Estimates

### Unconditional baseline impulse responses (key quantitative outcomes)
- Cumulative increase in real GDP per capita over an eight-year horizon:
  - Approximately 2 percent following a Government Effectiveness Index (GEE) shock.
  - About 4 percent following a Political Civil Liberties Index (PRCL) shock.
- State capacity shocks:
  - No statistically significant impact on real GDP per capita growth at any horizon (confidence intervals consistently encompass zero).
- Asymmetry:
  - Institutional shocks (GEE, PRCL) generate strong and persistent positive effects on growth; state capacity shocks do not.

### Heterogeneity by income group
- Advanced Economies (AEs):
  - State capacity shock: borderline statistically significant negative impact on growth over the medium term.
  - GEE shocks: positive and statistically significant effect on growth.
  - PRCL shocks: no statistically significant effect.
- Emerging Market and Developing Economies (EMDEs):
  - State capacity shocks: no statistically significant effect on growth.
  - GEE shocks: positive and borderline statistically significant impact.
  - PRCL shocks: positive and statistically significant impact.
- Asymmetry of positive vs negative governance shocks:
  - EMDEs: positive GEE and PRCL shocks lead to statistically significant GDP increases; negative PRCL shocks can yield significant losses.
  - AEs: GEE positive shocks boost GDP; PRCL effects muted.

---

### 6.2 Nonlinear LP Results — Complementarity and State-Dependence

### Core conditional findings
- Complementarity:
  - IN reforms (especially GEE) are most effective when state capacity is high.
  - SC upgrades are more effective when implemented in environments with credible institutions.
- Advanced Economies (conditional effects):
  - SC shock: negative and significant effects on GDP when GEE is low; positive and significant when GEE is high.
  - GEE shock: positive and significant effects only when SC is high.
  - PRCL shocks: do not display significant differences between high- and low-capacity settings.
- Emerging Markets and Developing Economies:
  - SC shocks: do not produce significantly different effects across IN levels.
  - GEE shocks: growth benefits materialize only when SC is high.
  - PRCL shocks: do not show significant differences based on SC level; PRCL effects may depend more on political context or regime stability.
- Policy synthesis:
  - Neither state capacity nor institutional quality alone ensures sustained growth; sequencing and coordination matter—state capacity (engine) and institutional quality (steering) must advance together.

---

### 6.3 Transmission Channels

### GDP components
- Advanced Economies:
  - Investment: negative and statistically significant response to SC shocks in the medium term.
  - GEE shocks: positive and significant effect on investment.
  - Private consumption: positive but borderline significant response to GEE shocks.
  - Public consumption: positive and significant response to both SC and GEE shocks.
  - Exports and imports: borderline significant declines following SC shocks.
- Emerging Markets and Developing Economies:
  - Private consumption: positive (borderline significant) response to GEE shocks.
  - Exports: rise significantly in response to both GEE and PRCL shocks.
  - Imports: positive and significant increase following SC shocks.

### Labor market
- Across AEs and EMDEs:
  - Shocks to SC, GEE, and PRCL show no statistically significant impact on employment or labor force participation (LFP), except:
    - In EMDEs, PRCL shocks are associated with a positive and significant effect on LFP.
  - SC shocks exhibit negative but statistically insignificant effects on employment and LFP.

### Economic stability
- Advanced Economies:
  - SC shocks: statistically significant reduction in CPI inflation and improvement in the overall fiscal deficit.
  - GEE shocks: positive and borderline significant effect on the overall fiscal balance.
- EMDEs:
  - No statistically significant effects on inflation or fiscal balance observed.

---

### Diagnostics, Robustness, and Key Statistics

### Pre-trend and exogeneity diagnostics
- Appendix Table A6 and Figure A6:
  - Coefficients on lagged growth are uniformly small and generally statistically insignificant.
  - Pre-trends Local Projection Test: forward-looking output dynamics do not systematically predict timing of institutional shocks; coefficients on future GDP growth are small and generally statistically indistinguishable from zero.
  - Regressions of shocks on lagged macro controls mostly show R-squared values close to zero; select significant small coefficients noted (e.g., PRCL L1 coefficient -0.002, P>|t| 0.000).

### IV evidence
- IV findings broadly confirm baseline results.
- For GEE shocks, IV estimates are statistically larger than baseline, suggesting attenuation bias in baseline estimates.

### Annex I — Descriptive statistics (selected exact values)
- State Capacity:
  - Obs.: 2048; Mean: 0.528; S.D.: 0.943; Min: -2.31; Max: 2.96.
- Government Effectiveness:
  - Obs.: 3178; Mean: 0.497; S.D.: 0.200; Min: 0.000; Max: 1.000.
- Political Civil Liberties:
  - Obs.: 3993; Mean: 0.692; S.D.: 0.276; Min: 0.024; Max: 0.985.
- Inflation rate (annual percentage change in CPI):
  - Obs.: 3109; Mean: 0.016; S.D.: 0.055; Min: -0.802; Max: 0.677.
- Real GDP per capita growth (annual growth rate):
  - Obs.: 2956; Mean: 0.049; S.D.: 0.065; Min: -0.184; Max: 1.267.
- Gross Fixed Capital Formation (GFCF) (%GDP):
  - Obs.: 2697; Mean: 24.355; S.D.: 8.377; Min: -3.945; Max: 76.782.
- Trade Openness (exports + imports as %GDP):
  - Obs.: 2790; Mean: 90.276; S.D.: 54.792; Min: 15.281; Max: 442.62.
- Labor force participation rate:
  - Obs.: 2972; Mean: 43.207; S.D.: 13.368; Min: 13.834; Max: 83.03.
- Overall Balance (%GDP):
  - Obs.: 2450; Mean: -1.8542; S.D.: 4.549; Min: -31.312.

---

### Policy recommendations and limitations

### Policy-relevant recommendations (preserved)
- Target Complementarity:
  - Reforms are most effective when SC and IN are strengthened together; IN reforms (especially GEE) offer high value when paired with sufficient administrative and implementation capacity.
- Sequencing:
  - In weak institutional environments, basic rule-of-law reforms may need to precede or accompany administrative overhauls.
- Country-specific tailoring:
  - Tailor governance reform strategies to development context: AEs should emphasize incremental efficiency-enhancing reforms; EMEs should prioritize core administrative capacity and institutional stabilization; LICs should adopt comprehensive joint reforms (anti-corruption, rule of law, public service delivery).
- Stakeholder engagement:
  - Engage civil society, the private sector, and external partners; multilateral institutions can provide tailored technical assistance and sustained capacity-building.

### Limitations and avenues for future research
- Limitations:
  - SC measures do not capture fiscal capacity explicitly.
  - LP methods are well-suited to short- and medium-term dynamics but may not fully capture long-run effects or structural persistence.
- Suggested extensions:
  - Incorporate explicit fiscal capacity indicators.
  - Explore sectoral heterogeneity.
  - Study interactions between governance and external shocks (climate change, demographic pressures, global financial cycles).

*International Monetary Fund — wpiea2025117-print-pdf (Section 4.1 and related chapters as provided).*

### 4.1 Identifying Shocks to State Capacity and Institutions ..............................................................

### 4.1 Identifying Shocks to State Capacity and Institutions

### Section location and context
- Located on page 13 of the PDF.
- Appears as the first subsection within Part IV, preceding "4.2 Local Projections" (page 14).

### Structural role in the report
- Introduces the identification of shocks to state capacity and institutions that are subsequently analyzed using local projections in section 4.2.
- Precedes the Data and Stylized Facts (V), Results and Discussion (VI), and Conclusion and Policy Implications (VII) sections.

*Source: wpiea2025117-print-pdf - 4.1 Identifying Shocks to State Capacity and Institutions (page 13).*

### References .............................................................................................................

### wpiea2025117-print-pdf - References .............................................................................................................

### I. Introduction — framing, concepts, and data
- Core concepts and distinctions
  - Institutions: "Formal and informal rules, norms, and organizations that structure societal interactions." (table language preserved).
  - State capacity (SC): "The ability of a government to implement policies, enforce laws, and provide public goods and services." (table language preserved).
  - State capacity is emphasized as quality/effectiveness rather than size of government.
- Historical and theoretical context (preserved references as presented)
  - Solow (1956) on growth as capital/technology; North (1990), North and Thomas (1973), North and Weingast (1989) on institutions as "rules of the game."
  - Developmental state literature: Johnson (1982), Amsden (1989), Wade (1990), Skocpol (1985).
  - State capacity conceptualizations: Mann (1986) "infrastructural power"; Berwick and Christia (2018) extraction/coordination/compliance; Besley et al. (2009) bureaucratic quality.
  - Formal work linking state capacity to growth: Besley and Persson (2010); modern empirical expansion (Dinecco and Wang, 2024).
- Empirical approach and data summary
  - Sample: up to 130 countries from 1970 to 2022 (with restrictions noted below).
  - Estimation method: local projections (LP) approach of Jordà (2005).
  - Operationalization:
    - State capacity (SC): Hanson and Sigman (2021) state capacity index, available through 2015.
    - Institutional quality (IN): V-Dem indicators Government Effectiveness (GEE) and Political and Civil Liberties (PRCL), available through 2022.
  - Sample restriction for combined-specification consistency: sample restricted to end in 2015.
- Shock construction and scaling
  - Two shock types:
    - Residual-based shocks: isolate large, unexpected changes after controlling for autoregressive dynamics and global trends.
    - Percentile-based (differenced) shocks: unusually large annual improvements in the top decile of a country’s own historical distribution.
  - Continuous shocks: LP coefficients scaled by the average magnitude of the governance change to express results in terms of a typical large institutional shift.
- Identification / instruments
  - SC shocks instrumented with natural disasters with high fatality rates as exogenous triggers of state-building episodes.
  - IN shocks (GEE and PRCL) instrumented with average annual changes in Polity2 scores among countries within the same IMF income group (excluding the target country), exploiting regional institutional diffusion and democratization momentum.
- Four stated contributions of the paper (verbatim structure preserved)
  - Empirical distinction between roles of state capacity and institutional quality.
  - Short- to medium-term effects of major governance improvements capturing transitional costs and early benefits.
  - Addressing endogeneity with external instruments (natural disasters; Polity2-based instrument).
  - Examination of nonlinearities and complementarity between SC and IN.

### II. Dissecting the interaction between institutional quality and state capacity — mechanisms and empirical checks
- Theoretical mechanisms and bidirectionality
  - Institutions (transparency, reduced corruption, accountability) can strengthen SC via enforcement of property rights and contract law.
  - High-capacity states (efficient tax systems, robust legal enforcement, reliable public service delivery) can bolster institutions by supporting regulation, reducing uncertainty, and upholding law and order.
  - The two can create a virtuous cycle; synergy yields positive correlation (Evans and Rauch, 1999; Rodrik, 2004).
- Multicollinearity assessment in empirical specifications
  - For raw variables: mean VIF remains below 6.
  - For shock-based LP specifications: mean VIF is close to 1, suggesting virtually no multicollinearity for shocks.
  - Strategy: analyze SC and IN separately and jointly; include interaction terms in select specifications to capture interplay without inflating multicollinearity risks.
- Enforcement vs. legitimacy
  - State capacity enforces rules; institutions lend legitimacy, increasing compliance and reducing resistance to reform (Besley and Persson, 2010; North, 1990).
  - Strong institutions curb corruption and undue influence, promoting equitable implementation of reforms (Mauro, 1995).
- Empirical patterns and heterogeneity (cases summarized as in text)
  - Positive co-movement examples: Republic of Korea post-Korean War (state-led development followed by democratization and institutional improvement).
  - Divergent examples: resource-rich autocracies (e.g., Equatorial Guinea) with revenue-related capacity but poor institutional quality; fragile/conflict-affected states with low/uncorrelated SC and IN; Venezuela and Zimbabwe as cases of misalignment leading to instability or collapse.
- Figure 1 (description preserved)
  - Presents unconditional bivariate correlations between normalized state capacity, institutions, and real GDP pc (1970-2015).
  - Note: "Unconditional bivariate correlations. Shaded grey areas denote 95 percent confidence with the blue continuous line denoting the linear fit."

### III. Relative importance of institutions vs. state capacity across development levels — complementarities, thresholds, and nonlinearities
- Cross-conditional effects and complementarities (verbatim logic preserved)
  - When SC and IN are both robust: mutual reinforcement creates a stable foundation for growth via improved public goods, property rights enforcement, and efficient administration.
  - If one component is stronger than the other:
    - Institutional improvements require SC to enforce reforms; weak SC undermines institutional gains (Besley and Persson, 2010; Evans and Rauch, 1999).
    - SC improvements require strong institutions to direct capacity toward productive ends; weak institutions risk rent-seeking/corruption (Mauro, 1995; Tanzi and Davoodi, 1997).
- Thresholds and diminishing returns
  - Potential threshold effects: benefits of SC improvements may be minimal until institutions reach a certain effectiveness threshold; vice versa for institutional reforms needing administrative support.
  - Diminishing returns can occur when either SC or IN is very high relative to the other.
- Empirical strategy for nonlinearities
  - Use of a Smooth Transition Autoregressive (STAR) model adapted to the context of nonlinear LPs to capture threshold effects, interactions, and diminishing returns (method stated exactly).
- Key empirical messages (as stated in paper summary)
  - Shocks to both SC and IN have significant and persistent positive effects on real GDP per capita growth, particularly in emerging and developing economies.
  - GEE and PRCL improvements are especially powerful drivers of growth.
  - SC shocks tend to be more effective when supported by strong institutions.
  - Robust complementarity: improvements in one governance dimension magnify the effect of the other.
  - Robustness: results hold across alternative shock definitions, model specifications, and data sources.
  - IV evidence: IV estimates are generally consistent with baseline findings; for GEE shocks IV estimates are statistically larger than baseline, suggesting attenuation bias may understate their full impact.
- Policy-relevant implications embedded in the analysis
  - Institutional reforms enhancing government effectiveness offer high value when paired with sufficient administrative and implementation capacity.
  - Tailoring governance reform strategies to different development contexts is recommended given cross-conditionalities and threshold effects.

*Source: Excerpted content from wpiea2025117-print-pdf - References (IMF), sections I–III as provided.*

### Section 4.1). By allowing for nonlinear dynamics, we are able to identify how the growth effects

### Section 4.1 Identifying Shocks to State Capacity and Institutions

### Nonlinear dynamics and cross-conditionalities
- Growth effects of improvements in state capacity (SC) and institutional quality (IN) vary depending on initial conditions and relative levels of governance.
- In highly developed institutional environments, further increases in state capacity might yield limited additional growth if institutions are already performing optimally.
- In countries with high state capacity, further institutional reform might yield diminishing returns without corresponding enhancements in state capacity to administer reforms effectively.
- Catalytic effects: a major change in one dimension (e.g., institutional reform) can initiate or accelerate improvements in the other, creating a virtuous cycle of growth.
- Example: Rwanda — strong state capacity post-1994 genocide facilitated gradual establishment of institutional frameworks, enabling better governance, resource allocation, and public service delivery; illustrates how significant shocks can reset governance trajectories and foster virtuous cycles when initial state capacity is leveraged.

### Identification challenge
- SC and IN indicators are highly persistent and evolve slowly, reflecting structural factors (development levels, demographic trends, institutional legacies).
- Simple year-on-year changes risk capturing routine endogenous dynamics rather than discrete, plausibly exogenous shifts.
- The identification strategy aims to isolate large and unexpected deviations that are more plausibly exogenous to contemporaneous macroeconomic conditions to improve causal inference.

### Multi-step procedure to generate shock series
- Step 1: Model expected evolution of indicators with autoregressive models augmented with macroeconomic fundamentals. For each country, estimate:
  - 푙표푔(퐺푂푉푖푡)=훼푖+훽1푙표푔(퐺푂푉푖푡−1)+훽2푙표푔(퐺퐷푃푖푡)+훽3푙표푔(푃푂푃푖푡)+푢푖푡
  - where 퐺푂푉푖푡 represents the relevant indicator (state capacity or institutional quality), 퐺퐷푃푖푡 and 푃푂푃푖푡 denote real GDP per capita and population, respectively.

- Step 2: Extract residuals 푢푖푡, representing deviations from the expected path that capture unanticipated movements (political reforms, institutional crises, external shocks) and are more likely to be exogenous with respect to near-term macroeconomic outcomes.

- Step 3: Transform continuous residuals into discrete shock indicators using a country-specific threshold:
  - 푆ℎ표푐푘푖푡=
    { 1 if 푢푖푡>푢̅푖+휎푖
      0 otherwise
  - where 푢̅푖 and 휎푖 denote the mean and standard deviation of the residuals for country i.
  - Threshold: residual exceeds its historical country-specific mean by more than one standard deviation.
  - Purpose: filter out minor fluctuations and isolate large, abrupt deviations more likely to reflect genuine shocks rather than measurement noise or normal cyclical variation.

### Alternative percentile-based shock definition
- Construct an alternative set of shocks based on the annual change (first difference) of the relevant indicators.
- Define a shock as any year when the annual change ranks among the top 10 percent of all observed changes in that country.
- Operationalization: for each country, calculate the 90th percentile of the annual change distribution and classify a year as a shock if the change exceeds this threshold.
- Rationale: better capture relative shifts within each country’s historical experience for slow-moving institutional variables; ensures shocks reflect unusually large movements relative to each country's typical institutional dynamics.

### Empirical strengths and implications
- Dual strategy captures both:
  - Large residual-driven deviations from predicted institutional dynamics.
  - Relatively rare but sizable annual changes in the relevant indicators.
- Improves robustness and interpretability of empirical analysis.
- Aligns with literature emphasizing macroeconomic relevance of large institutional shifts.

*International Monetary Fund — wpiea2025117-print-pdf (Section 4.1).*

### 4.2 Local Projections

### 4.2 Local Projections

### Rationale for using Local Projections (LP) over VAR
- LP (Jordà, 2005) is used to estimate impulse-response functions as a flexible alternative to VAR, supported by Auerbach and Gorodnichenko (2013) and Romer and Romer (2019).
- Advantages in the panel setting:
  - Shocks (state capacity and institutions) are initially assumed orthogonal to contemporaneous and expected future macroeconomic conditions; endogeneity is addressed via IV robustness checks.
  - LP reduces the number of estimated parameters relative to VARs, avoiding the need to estimate equations for other dependent variables; beneficial in finite samples and for shorter horizons.
  - LP is well suited to estimating nonlinearities and interaction effects; easier to implement when nonlinearity complicates VAR-to-VMA inversion (Plagborg-Møller and Wolf, 2021).
  - Error-term cross-country correlation is handled via clustering or Driscoll-Kraay (1998) standard errors; SCC standard errors are used following Herbst and Johannesen (2024).

### Linear LP model specification and estimation details
- Basic linear LP regression (notation preserved from source):
  - ∆log푦푖,푡+ℎ = 훼푖,ℎ + 훿푡,ℎ + ∑훽푗,ℎ 푑푖,푡−푗2 푗=0 + ∑훽푙,ℎ (푙표푔푦푖,푡−푙 − 푙표푔푦푖,푡−1−푙)1 푙=0 + ∑훽ℎ 푑푖,푡+ℎ ℎ ℎ=1 + ∑훽푐,ℎ ′ 푋푖,푡−푐1 푐=0 + 푢푖,푡+ℎ  (equation (3))
- Key parameterization and sample facts:
  - Forecast horizon h = 1, ..., 8.
  - ∆log푦푖,푡+ℎ = log푦푖,푡+ℎ − log푦푖,푡 (real GDP per capita cumulative growth over horizon).
  - 훼푖 are country fixed effects; 훿푡 are time fixed effects.
  - 푑푖,푡 denotes state capacity (SC) or institutions (IN) shocks, defined as large, unexpected changes (residuals exceeding one standard deviation); IRFs are scaled by the average change in the relevant indicator observed when shocks occur.
  - When forecasting one year ahead, maximum of 3826 observations for state capacity changes; for each additional year-ahead forecast, lose up to 129 observations (number of cross-sectional units).
  - For institutions: 1126 observations corresponding to 55 countries and 4719 observations corresponding to 130 countries for the main proxies (World Bank GEE and v-Dem Political Civil Liberties Index, respectively).
- Lag selection and stationarity:
  - Treatment lags: AIC indicates 2 lags of the treatment variable.
  - Lags of ∆log푦푖,푡: AIC indicates 1 lag.
  - Data is stationary as |∑ 훽푙,ℎ|<1 4 푙=0 in all specifications.
  - ∑ 훽푙,ℎ 4 푙=0 is between small for all h, indicating low persistence in estimated models.
  - Time dimension t = 46 for most cross-sectional units; bias from short t is unlikely per Herbst and Johannesen (2024).
- Overlapping-horizon bias correction:
  - Include leads ∑훽ℎ 푑푖,푡+ℎ ℎ ℎ=1 to capture Teulings and Zubanov (2014) correction; leads are statistically significant for most combinations of 푦푖,푡 and h.
- Standard errors and cross-sectional dependence:
  - Use Spatial Correlation Consistent (SCC) standard errors (Driscoll and Kraay, 1998).
  - Pesaran (2015) test used to detect cross-sectional dependence; test is often significant.
  - SCC standard errors are cluster robust and robust to spatial correlation.

### Control variables and theoretical links
- Control vector 푋푖,푡 includes contemporaneous and first lags of:
  - Investment (Gross Capital Formation): linked to Solow (1956) and Barro (1991).
  - Inflation Rate: related to Fischer (1993), Barro (1995), Bruno and Easterly (1998).
  - Trade Openness: linked to Frankel and Romer (1999), Sachs and Warner (1995).
  - Population Growth: linked to Bloom and Williamson (1998), Barro and Becker (1989).
  - Policy Uncertainty: linked to Baker, Bloom, and Davis (2016); Alesina and others (1996).

### State-dependent (nonlinear) LPs and STAR augmentation
- Motivation:
  - To test whether effects of SC changes on growth depend on prevailing institutional quality and vice versa.
  - Recognizes potential interdependence between state capacity and institutional quality; policy interventions should consider both jointly.
- Augmented specification (equation (4), preserved notation):
  - ∆log푦푖,푡+ℎ = 훼푖,ℎ + 훿푡,ℎ + ∑훽푗,ℎ 퐿 퐹(푧푖,푡) 푑푖,푡−푗2 푗=0 + ∑훽푗,ℎ 퐻 퐹(1−(푧푖,푡)) 푑푖,푡−푗1 푗=0 + ... + 푢푖,푡+ℎ
  - Weighting function: 퐹(푧푖푡) = exp(−훾푧푖푡) / (1+exp(−훾푧푖푡)), 훾>0; 푧푖푡 is an indicator of IN or SC normalized to zero mean and unit variance.
  - 훽푗=0,ℎ 퐿 and 훽푗=0,ℎ 퐻 construct IRFs for low and high regimes of the conditioning variable (퐹(푧푖푡) ≈1 for low; 1−퐹(푧푖푡) ≈1 for high).
- Advantages of LPs in state-dependent context:
  - Simpler to account for state-dependence in a panel than VARs or regime-switching VARs.
  - Do not require assumptions on duration of states or transition mechanisms.
- Caveat:
  - State should be uncorrelated with the shock; if state depends on the shock, LPs recover responses only to small shocks (Gonçalves and others, 2024).
  - In this estimation sample, unconditional correlation between state capacity shocks and institutional shocks (World Bank GEE or v-Dem Political Civil Liberties Index) is very low and statistically insignificant at the 5 percent level.

### Data, measurement, and stylized facts
- State capacity measure:
  - Proxy integrates 21 indicators across extractive, coercive, and administrative capacity (Hanson and Sigman, 2021).
  - Dataset spans 177 countries, years 1960–2015; higher values indicate greater state capacity.
  - Method: latent variable modelling.
- Institutional quality measures:
  - Government Effectiveness Index (GEE), World Bank WGI: captures quality of public services, competence of civil service, policy formulation/implementation; covers over 200 countries since the 1990s.
  - Political Civil Liberties Index (central estimate), V-Dem (2025 edition), processed by Our World in Data: spans 1789–2024; captures freedoms of expression and association.
- Complementarity:
  - GEE emphasizes administrative/managerial dimensions; V-Dem Political Civil Liberties emphasizes democratic principles and civil rights—together provide a more complete assessment of institutional quality.
- Sample visual and descriptive findings (preserved descriptions):
  - When forecasting one year ahead: maximum of 3826 observations for state capacity changes.
  - Figures (described in text):
    - Figure 2.a: State Capacity Index, 2015 (normalized).
    - Figure 2.b: Government Effectiveness Rating, 2015 (normalized).
    - Figure 3: Box-Whisker of normalized (0-1) variables by income group (State Capacity, 2015; Government Effectiveness, 2015). Note: Blue = AE, Red = EMEs, Green = LICs.
    - Figures 4.a and 4.b: Interquartile range of normalized (0-1) state capacity and government effectiveness index by income group, 1970-2015. Note: Red = top quartile, Blue = median; Green = bottom quartile.
  - Stylized patterns:
    - Clear correlation between stage of development and higher values in Europe, Australia, and North America.
    - State capacity shows an upward trend across all income groups.
    - Government effectiveness appears more stable over time with no clear upward or downward pattern.
- Robustness and tests:
  - Fisher-type panel stationarity tests confirm stationarity (available on request).
  - Robustness check using bias-corrected panel estimator derived in Herbst and Johannesen (2024) yields very similar results to eq.(1).
  - Pesaran (2015) cross-sectional dependence test often significant; SCC standard errors mitigate spatial correlation concerns.
  - Results on some tests and additional robustness checks are available on request.

*Source: IMF working paper chapter “4.2 Local Projections.”*

### 6.1 Baseline LP results

### 6.1 Baseline LP results

### Unconditional baseline impulse responses
- Method: Local projections plotting dynamic responses to positive, unanticipated shocks in state capacity and institutional quality. Solid blue lines show estimated impulse responses; dark and light grey shaded areas denote the 90 percent and 95 percent confidence intervals, respectively.
- Key quantitative outcomes:
  - Cumulative increase in real GDP per capita over an eight-year horizon:
    - Approximately 2 percent following a Government Effectiveness Index (GEE) shock.
    - About 4 percent following a Political Civil Liberties Index (PRCL) shock.
  - Positive state capacity shocks: no statistically significant impact on real GDP per capita growth at any horizon (confidence intervals consistently encompass zero).
- Asymmetry: institutional shocks (GEE, PRCL) generate strong and persistent positive effects on economic growth; state capacity shocks do not.

### Robustness to dynamic heterogeneity
- Concern: heterogeneous dynamic responses across countries could bias panel impulse responses.
- Approach: re-estimated responses using simple and weighted averages of country-specific estimates (following Pesaran, 2006; Canova, 2024).
- Result: re-estimated results (not shown) remain closely aligned with panel estimates, suggesting minimal bias from dynamic heterogeneity.

### Heterogeneity by income group (AEs vs EMDEs)
- Advanced Economies (AEs):
  - State capacity shock: associated with a borderline statistically significant negative impact on growth, especially over the medium term.
  - GEE shocks: produce a positive and statistically significant effect on growth.
  - PRCL shocks: no statistically significant effect on growth.
  - Interpretation: diminishing marginal returns to state-building reforms in high-capacity states; bureaucratic restructuring or administrative frictions may temporarily suppress growth.
- Emerging Market and Developing Economies (EMDEs):
  - State capacity shocks: no statistically significant effect on growth.
  - GEE shocks: yield a positive and borderline statistically significant impact.
  - PRCL shocks: produce a positive and statistically significant impact on real GDP per capita.
  - Interpretation: lower baseline institutional quality implies that modest improvements (especially in PRCL and GEE) can trigger substantial economic gains; implementation constraints may mute state capacity shocks.

### Asymmetry of positive vs negative governance shocks
- Importance: positive and negative changes operate through distinct mechanisms; Appendix Figure A1 contrasts positive and negative shocks.
- Advanced Economies:
  - State capacity: no statistically significant difference between positive and negative shocks (effects small and statistically muted).
  - GEE: positive shocks significantly boost GDP; negative shocks are economically negative but statistically insignificant.
  - PRCL: no significant asymmetry observed.
- EMDEs:
  - State capacity: positive or negative shocks do not produce statistically significant growth effects.
  - GEE and PRCL: positive shocks lead to statistically significant increases in GDP; negative shocks either yield significant losses (PRCL) or are directionally negative but statistically insignificant (GEE).
  - Interpretation: outsized returns to institutional improvement in lower-capacity settings; negative institutional changes can be economically damaging in EMDEs.

---

### 6.2 Nonlinear LP results

### Interaction and complementarity between state capacity and institutional quality
- General finding: effects of governance shocks are conditional, asymmetric, and reflect deep complementarities between state capacity and institutional quality.
- Theoretical framing: supports perspectives that economic development depends on mutual reinforcement of administrative capacity and institutional quality.

### Advanced Economies (conditional effects)
- State capacity shock:
  - Generates statistically significant negative effects on GDP when institutional quality (GEE) is low.
  - Generates positive and significant effects when GEE is high.
  - Interpretation: capacity expansion without strong institutions can fuel inefficiencies, red tape, or rent-seeking; benefits realized when institutions can guide and constrain capacity.
- GEE shocks:
  - Positive and statistically significant effects only when state capacity is high.
  - Interpretation: institutional reforms require implementation infrastructure (personnel, systems, fiscal space) to be effective.
- PRCL shocks:
  - Do not display significant differences between high- and low-capacity settings.

### Emerging Markets and Developing Economies (conditional effects)
- State capacity shocks:
  - Do not produce significantly different effects across levels of institutional quality (possible reasons: structural rigidities, implementation delays, political resistance).
- GEE shocks:
  - Growth benefits materialize only when state capacity is high (replicates AE pattern).
- PRCL shocks:
  - Do not show significant differences based on the level of state capacity; economic relevance may depend more on political context or regime stability.
- Additional quantitative note: while there is slightly more uncertainty in EMDE estimates, medium-term effects (from year 6 onward) are positive and statistically significant.

### Policy-relevant synthesis
- Central lesson: neither state capacity nor institutional quality alone ensures sustained growth; they must progress together.
- Practical implication: one-dimensional reform risks limited or counterproductive outcomes; sequencing and coordination are critical—state capacity (engine) and institutional quality (steering) must advance in tandem to drive growth.

*Source: wpiea2025117-print-pdf - 6.1 Baseline LP results*

### 6.3 Channels

### 6.3 Channels

### GDP Components
- Disaggregation into private consumption, public consumption, investment (gross fixed capital formation), and net exports reveals heterogeneous effects by income group and governance dimension.
- Advanced Economies (AEs):
  - Investment exhibits a negative and statistically significant response to state capacity shocks in the medium term.
  - Institutional improvements (GEE shocks) have a positive and significant effect on investment.
  - Private consumption shows a positive but borderline significant response to GEE shocks.
  - Public consumption reacts positively and significantly to both state capacity and GEE shocks.
  - State capacity shocks lead to borderline significant declines in both exports and imports.
  - Interpretation: increasing state capacity without institutional reform may generate inefficiencies or regulatory overreach, deterring private investment; GEE improvements likely enhance regulatory quality and reduce policy uncertainty, encouraging capital formation.
- Emerging Markets and Developing Economies (EMDEs):
  - Private consumption responds positively to GEE shocks (borderline significant).
  - Exports rise significantly in response to both GEE and PRCL shocks.
  - State capacity shocks lead to a positive and significant increase in imports.
  - Interpretation: improvements in rule of law, transparency, and civil liberties (PRCL) reduce transaction costs and enhance export competitiveness; state capacity improvements likely improve customs administration, infrastructure, and domestic demand.

- Figure notes (as reported in the source):
  - The figure displays impulse responses of state capacity and institutional shocks on real GDP per capita growth.
  - Year 1 refers to the first year following the occurrence of the shock at Year 0.
  - Solid blue lines represent estimated responses; position at Year 8 indicates cumulative change in the log of real GDP per capita eight years after the shock.
  - Dark grey shaded areas denote 90% confidence intervals based on small-sample corrected (SCC) standard errors; light grey shaded areas indicate 95% confidence intervals.
  - Left panel: responses to a state capacity shock; right panels: responses to institutional shocks distinguishing between government effectiveness (GEE) and political civil liberties (PRCL).

### Labor Market
- Theoretical expectations:
  - A capable state supports job creation by enforcing labor laws, promoting workforce training, and providing a stable environment for business expansion.
  - Strong institutions reduce hiring frictions, foster fair labor practices, and encourage participation by lowering entry barriers and boosting market inclusiveness.
- Empirical results (Appendix Figure A2):
  - Across both AEs and EMDEs, shocks to state capacity, GEE, and PRCL show no statistically significant impact on employment or labor force participation (LFP).
  - Exception: in EMDEs, PRCL shocks are associated with a positive and significant effect on LFP.
  - State capacity shocks in both income groups consistently exhibit negative, albeit statistically insignificant, effects on employment and LFP.
  - Interpretation: expanded civil liberties can reduce discrimination and empower underrepresented groups to enter the formal workforce; transitional costs of expanding administrative reach (e.g., stricter enforcement or compliance burdens) may temporarily dampen labor market participation, particularly in the informal sector.

### Economic Stability
- Role of governance:
  - Effective governance—through strong state capacity and institutional quality—helps maintain low inflation and sound public finances, fostering an environment conducive to investment and long-term productivity.
- Advanced Economies (AEs):
  - State capacity shocks lead to a statistically significant reduction in both CPI inflation and the fiscal deficit (i.e., improve the overall balance).
  - GEE shocks are associated with a positive and borderline significant effect on the overall fiscal balance.
  - Interpretation: improvements in administrative capability contribute to macroeconomic stability via enhanced efficiency of public spending, strengthened tax collection, and reinforced fiscal control mechanisms.
- Emerging Markets and Developing Economies (EMDEs):
  - No statistically significant effects are observed for either inflation or fiscal balance.
  - Interpretation: structural challenges—higher exposure to external shocks, weaker monetary policy frameworks, or limited administrative capacity—may dampen the impact of governance improvements on macroeconomic outcomes.

*Source: IMF — 6.3 Channels*

### Appendix Table A6, the coefficients on lagged growth are uniformly small and, with few

### wpiea2025117-print-pdf - Appendix Table A6, the coefficients on lagged growth are uniformly small and, with few

### Diagnostic tests on lagged growth and pre-trends
- Appendix Table A6: the coefficients on lagged growth are uniformly small and, with few exceptions, statistically insignificant across all shock types.
- R-squared values: consistently close to zero, indicating that past output dynamics have limited explanatory power for the timing or occurrence of institutional shocks.
- Pre-Trends Local Projection Test:
  - Method: reverse local projection analysis regressing the probability of a governance shock on future real GDP growth, controlling for lagged fundamentals and fixed effects.
  - Results (Appendix Figure A6, Panels 1–3 for State Capacity, GEE, PRCL): estimated effects are small and generally statistically indistinguishable from zero, with confidence intervals consistently spanning the zero line.
  - Interpretation: forward-looking output dynamics do not systematically predict the timing of institutional shocks.
- Overall diagnostic conclusion: identified governance shocks are orthogonal to both lagged and anticipated economic conditions, supporting their treatment as plausibly exogenous events that can be credibly treated as causal drivers in the analysis of growth outcomes.

### Empirical design, data, and identification
- Sample and period: panel of up to 130 countries spanning 1970 to 2022.
- Governance measures:
  - State Capacity (SC): Hanson and Sigman (2021) index.
  - Institutional quality (IN): Government Effectiveness (GEE) from the World Bank’s Worldwide Governance Indicators and Political Rights and Civil Liberties (PRCL) from the V-Dem dataset.
- Shock construction:
  - Residual-based method: captures large unanticipated governance changes.
  - Percentile-based approach: identifies top-decile improvements in governance indicators within each country’s historical distribution.
- Estimation method: local projections (LP), incorporating nonlinear interactions and exploring multiple identification strategies.
- Instrumental variable (IV) strategy:
  - For SC shocks: natural disasters with high fatality rates used as plausibly exogenous triggers of state-building efforts.
  - For GEE and PRCL shocks: regional institutional diffusion instrumented with the average change in Polity2 scores among peer countries within the same income group (excluding the country in question).
  - IV findings: broadly confirm baseline results; for GEE, IV estimates are significantly larger than baseline, suggesting attenuation bias in baseline estimates.

### Main empirical findings
- Nonlinearities and complementarities:
  - Positive shocks to SC and IN have significant and persistent growth effects, particularly in low-income and emerging market economies.
  - Improvements in institutional quality, especially in government effectiveness (GEE), generate the strongest and most consistent gains.
  - SC shocks have more modest average effects but become substantially more powerful when implemented in environments with high rule of law, bureaucratic quality, and accountability.
  - Complementarity conclusion: IN reforms are most impactful in countries with sufficient administrative capacity; SC upgrades are more effective when supported by credible institutions.
- Heterogeneity by income group:
  - Low-income and emerging market economies: larger and more persistent growth gains from governance improvements.
  - Advanced economies: marginal governance improvements produce smaller effects and sometimes lead to short-term adjustment costs.

### Transmission mechanisms
- Investment channel:
  - Gross fixed capital formation increases significantly following governance improvements, particularly institutional reforms.
  - Emphasis: stable, predictable policy environments support capital accumulation.
- Macroeconomic stability:
  - SC reforms associated with improved fiscal performance, despite often involving increased public expenditure.
- Labor market:
  - SC shocks associated with temporary declines in labor force participation, possibly reflecting transitional disruptions during reform efforts.
- Inflation:
  - Inflation appears largely unaffected by governance shocks, consistent with monetary dynamics being shaped by broader policy regimes and external factors.

### Policy recommendations and actionable insights
- By country group:
  - Advanced Economies (AEs):
    - Focus on incremental governance improvements that reduce administrative frictions without creating institutional overreach.
    - Recommend: digitalization and simplification of regulatory frameworks to yield efficiency gains with minimal disruption.
  - Emerging Market Economies (EMEs):
    - Prioritize building core administrative capacity and improving policy effectiveness.
    - Recommend: institutional reforms to stabilize expectations and reduce legal and regulatory uncertainty to support investment.
  - Low-Income Countries (LICs):
    - Adopt comprehensive reforms that jointly enhance state capacity and institutional quality.
    - Priority areas: anti-corruption, rule of law, and public service delivery.
    - Regional notes: in Africa and other LIC contexts, governance improvements aligned with regional integration initiatives may amplify growth effects.
    - Role for multilateral institutions: support through tailored technical assistance and sustained capacity-building.
- General principles across contexts:
  - Target Complementarity: reforms most effective when SC and IN are strengthened together, particularly where one dimension is underdeveloped.
  - Focus on Sequencing: sequencing matters; in weak institutional environments, basic rule-of-law reforms may need to precede or accompany administrative overhauls.
  - Adapt to Local Contexts: reforms must be tailored to a country’s unique institutional, historical, and social landscape.
  - Engage Stakeholders: broad-based support is critical; engaging civil society, the private sector, and external partners can increase buy-in and reduce reform resistance.

### Limitations and avenues for future research
- Limitations:
  - SC measures do not capture fiscal capacity explicitly, limiting differentiation between administrative efficiency and resource mobilization.
  - Local projection (LP) methods are well-suited to short- and medium-term dynamics but may not fully capture long-run effects or structural persistence.
- Suggested extensions:
  - Incorporate explicit fiscal capacity indicators.
  - Explore sectoral heterogeneity.
  - Study interactions between governance and external shocks such as climate change, demographic pressures, or global financial cycles.

### Annex I — Descriptive statistics (Table A.1)
- State Capacity
  - Obs.: 2048
  - Mean: 0.528
  - S.D.: 0.943
  - Min: -2.31
  - Max: 2.96
  - Source: Hanson and Sigman (2021)
- Government Effectiveness
  - Obs.: 3178
  - Mean: 0.497
  - S.D.: 0.200
  - Min: 0.000
  - Max: 1.000
  - Source: World Bank Governance Indicators
- Political Civil Liberties
  - Obs.: 3993
  - Mean: 0.692
  - S.D.: 0.276
  - Min: 0.024
  - Max: 0.985
  - Source: V-Dem via Our World in Data
- Inflation rate (annual percentage change in CPI)
  - Obs.: 3109
  - Mean: 0.016
  - S.D.: 0.055
  - Min: -0.802
  - Max: 0.677
  - Source: IMF
- Real GDP per capita growth (annual growth rate)
  - Obs.: 2956
  - Mean: 0.049
  - S.D.: 0.065
  - Min: -0.184
  - Max: 1.267
  - Source: IMF
- Population growth (annual percentage change)
  - Obs.: 3146
  - Mean: 1.427
  - S.D.: 1.605
  - Min: -6.852
  - Max: 19.360
  - Source: World Bank World Development Indicators
- Gross Fixed Capital Formation (GFCF) (%GDP)
  - Obs.: 2697
  - Mean: 24.355
  - S.D.: 8.377
  - Min: -3.945
  - Max: 76.782
  - Source: World Bank World Development Indicators
- Trade Openness (exports + imports as %GDP)
  - Obs.: 2790
  - Mean: 90.276
  - S.D.: 54.792
  - Min: 15.281
  - Max: 442.62
  - Source: World Bank World Development Indicators
- Uncertainty Index
  - Obs.: 2364
  - Mean: 0.192
  - S.D.: 0.202
  - Min: 0
  - Max: 1.821
  - Source: Baker, Bloom, and Davis (2016)
- Log real GFCF
  - Obs.: 2462
  - Mean: 10.123
  - S.D.: 2.453
  - Min: 1.156
  - Max: 18.043
  - Source: World Bank World Development Indicators
- Log real Private Consumption
  - Obs.: 2447
  - Mean: 11.054
  - S.D.: 2.400
  - Min: 5.503
  - Max: 18.367
  - Source: World Bank World Development Indicators
- Log real Public Consumption
  - Obs.: 2448
  - Mean: 9.661
  - S.D.: 2.386
  - Min: 3.148
  - Max: 17.255
  - Source: World Bank World Development Indicators
- Log real Exports
  - Obs.: 2458
  - Mean: 7.541
  - S.D.: 1.858
  - Min: 2.625
  - Max: 12.275
  - Source: World Bank World Development Indicators
- Log real Imports
  - Obs.: 2458
  - Mean: 7.716
  - S.D.: 1.618
  - Min: 3.000
  - Max: 12.099
  - Source: World Bank World Development Indicators
- Log employment
  - Obs.: 2972
  - Mean: 15.284
  - S.D.: 1.816
  - Min: 10.917
  - Max: 20.644
  - Source: World Bank World Development Indicators
- Labor force participation rate
  - Obs.: 2972
  - Mean: 43.207
  - S.D.: 13.368
  - Min: 13.834
  - Max: 83.03
  - Source: World Bank World Development Indicators
- Overall Balance (%GDP)
  - Obs.: 2450
  - Mean: -1.8542
  - S.D.: 4.549
  - Min: -31.312

*International Monetary Fund — content unit: wpiea2025117-print-pdf*

### 19.087 IMF

### wpiea2025117-print-pdf - 19.087 IMF

### Unconditional local projections — state capacity shocks (Table A.2)
- Outcome: real GDP per capita growth (percent) at horizons gdp1 ... Gdp7.
- Immediate (shock1) point estimates:
  - shock1: -0.000, 0.000, -0.000, -0.000, -0.000, -0.000, -0.001 (b/se in parentheses across columns: (0.002) ... (0.002))
  - L.shock1: 0.001, 0.001, 0.001, 0.002, 0.002, 0.002, 0.002 (standard errors (0.002)...(0.003))
- Notable distributed-lag / forecast coefficients (Fk.shock1):
  - F2.shock1: -0.005**; -0.004**; -0.004**; -0.005**; -0.004**; -0.004** (all se (0.002))
  - F3.shock1: -0.003; -0.003; -0.004+; -0.004; -0.004 (se (0.003))
  - F6.shock1: 0.005**; 0.005** (se (0.002))
  - F7.shock1: -0.006* (se (0.003))
- Selected control-variable coefficients (example column patterns):
  - popgr: -0.016***, -0.017***, -0.016**, -0.017**, -0.009***, -0.009***, -0.008*** (se varying; e.g., first (0.006))
  - L.popgr: 0.016**, 0.017**, 0.016**, 0.015*, 0.007***, 0.007**, 0.006** (se (0.007)...(0.003))
  - open: 0.000**, 0.000*, 0.000*, 0.000**, 0.001***, 0.001***, 0.001*** (se (0.000))
  - L3.open: -0.001*** across columns (se (0.000))
  - c_gdp: 0.103, 0.094, 0.087, 0.117, 0.201***, 0.203***, 0.203*** (se reported)
- Sample and fit:
  - Observations: 3,826; 3,700; 3,575; 3,451; 3,328; 3,206; 3,085 (by horizon)
  - Number of groups: 129; 128; 126; 125; 124; 123; 122
  - r2_w: 0.134; 0.136; 0.133; 0.142; 0.162; 0.171; 0.174
- Notes: Estimates of eq. (3). Spatial correlation consistent standard errors in parentheses. Significance markers: * p < 0.10, ** p < 0.05, *** p < 0.01.

### Unconditional local projections — government effectiveness (GEE) shocks (Table A.3)
- Immediate (shock2) point estimates:
  - shock2: 0.005**; 0.005**; 0.006**; 0.006**; 0.007**; 0.007**; 0.006** (se (0.002)...(0.003))
  - L.shock2: 0.002; 0.002*; 0.002+; 0.002+; 0.003**; 0.004***; 0.003** (se (0.001))
- Notable distributed-lag / forecast coefficients (Fk.shock2):
  - F.shock2: 0.003*; 0.004**; 0.004**; 0.004**; 0.004**; 0.004***; 0.003* (se (0.002))
  - F2.shock2: 0.005**; 0.006**; 0.006***; 0.006***; 0.007***; 0.006** (se (0.002))
  - F3.shock2: 0.002; 0.003**; 0.004**; 0.005***; 0.004** (se (0.002) or (0.001))
  - F6.shock2: 0.006**; 0.006*** (se (0.002))
  - F7.shock2: -0.003 (se (0.003))
- Selected control-variable highlights:
  - L2.lncpi: 0.050***, 0.052***, 0.051***, 0.055***, 0.054***, 0.051***, 0.047*** (se (0.010)...(0.013))
  - popgr: -0.006*** across horizons (se (0.001))
  - L.popgr: 0.006** ... 0.008*** ... 0.007** (se (0.002)-(0.003))
  - c_gdp: 0.221***, 0.216***, 0.207***, 0.190***, 0.212***, 0.186***, 0.171*** (se reported)
- Sample and fit:
  - Observations: 1,126; 1,072; 1,017; 962; 907; 852; 797
  - Number of groups: 55 (all horizons)
  - r2_w: 0.443; 0.448; 0.450; 0.460; 0.472; 0.489; 0.494

### Unconditional local projections — political rights and civil liberties (PRCL) shocks (Table A.4)
- Immediate (shock3) point estimates:
  - shock3: 0.001; 0.000; -0.001; -0.001; -0.001; -0.001; -0.001 (se (0.004)...(0.005))
  - L.shock3: -0.005*; -0.006**; -0.006*; -0.006*; -0.006*; -0.007*; -0.007* (se (0.003)...(0.004))
- Notable distributed-lag / forecast coefficients (Fk.shock3):
  - F.shock3: -0.006*; -0.005; -0.005; -0.005; -0.005; -0.005; -0.005 (se (0.003)-(0.004))
  - F2.shock3: -0.008***; -0.009***; -0.009***; -0.010***; -0.010***; -0.010*** (se (0.003))
  - F5.shock3: -0.008***; -0.008***; -0.008** (se (0.003))
  - F6.shock3: -0.003*; -0.002 (se (0.002))
- Selected control-variable highlights:
  - L3.lncpi: -0.009*; -0.008*; -0.009*; -0.009*; -0.009+; -0.008+; -0.008+ (se (0.005))
  - popgr: -0.013** ... -0.015** across horizons (se (0.005)-(0.006))
  - L.popgr: 0.013** ... 0.014** consistently (se (0.006)-(0.007))
  - open: 0.000* ... 0.000** (se (0.000))
  - L3.open: -0.000*** ... -0.001*** (se (0.000))
- Sample and fit:
  - Observations: 4,719; 4,602; 4,483; 4,360; 4,236; 4,112; 3,985
  - Number of groups: 130 (all horizons)
  - r2_w: 0.181; 0.186; 0.182; 0.132; 0.134; 0.134; 0.137

### Panel Granger causality tests (Table A5)
- Tests use Dumitrescu and Hurlin (2012) methodology; null: GDP growth does not Granger-cause shocks.
- State capacity (SC):
  - Lag order 1: W-bar 1.8197; Z-bar 5.5592; p-value (Z-bar) 0.0000; Z-bar tilde 4.7482; p-value (Z-bar tilde) 0.0000
  - Lag order 2: W-bar 2.9975; Z-bar 4.7840; p-value (Z-bar) 0.0000; Z-bar tilde 3.7842; p-value (Z-bar tilde) 0.0002
- GEE:
  - Lag order 1: W-bar 0.9548; Z-bar -0.2749; p-value (Z-bar) 0.7834; Z-bar tilde -0.8091; p-value (Z-bar tilde) 0.4185
  - Lag order 2: W-bar 1.9046; Z-bar -0.4102; p-value (Z-bar) 0.6816; Z-bar tilde -1.2421; p-value (Z-bar tilde) 0.2142
- PRCL:
  - Lag order 1: W-bar 1.4245; Z-bar 3.0020; p-value (Z-bar) 0.0027; Z-bar tilde 2.4880; p-value (Z-bar tilde) 0.0128
  - Lag order 2: W-bar 2.2229; Z-bar 1.1147; p-value (Z-bar) 0.2650; Z-bar tilde 0.5811; p-value (Z-bar tilde) 0.5612

### Regressions of shocks on lagged macro controls (Table A6)
- Purpose: test whether the occurrence of identified shocks is systematically predicted by recent macroeconomic conditions (Pooled OLS; one lag at a time).
- Selected results (Shock / Lag / Obs / R-squared / Coefficient / Std. Err. / t / P>|t| / 95% CI lower / upper):
  - SC L1: Obs 5,773; R-squared 0.000; Coefficient -0.001; Std. Err. 0.001; t -1.110; P>|t| 0.266; 95% CI [-0.002, 0.001]
  - SC L2: Obs 5,632; R-squared 0.002; Coefficient -0.002; Std. Err. 0.001; t -3.560; P>|t| 0.000; 95% CI [-0.004, -0.001]
  - SC L3: Obs 5,489; R-squared 0.000; Coefficient 0.000; Std. Err. 0.001; t 0.310; P>|t| 0.757; 95% CI [-0.001, 0.002]
  - GEE L1: Obs 1,833; R-squared 0.001; Coefficient 0.002; Std. Err. 0.002; t 0.990; P>|t| 0.322; 95% CI [-0.002, 0.007]
  - GEE L2: Obs 1,824; R-squared 0.000; Coefficient 0.001; Std. Err. 0.002; t 0.330; P>|t| 0.743; 95% CI [-0.004, 0.005]
  - GEE L3: Obs 1,815; R-squared 0.001; Coefficient 0.004; Std. Err. 0.002; t 1.590; P>|t| 0.112; 95% CI [-0.001, 0.008]
  - PRCL L1: Obs 7,345; R-squared 0.002; Coefficient -0.002; Std. Err. 0.001; t -3.950; P>|t| 0.000; 95% CI [-0.003, -0.001]
  - PRCL L2: Obs 7,178; R-squared 0.000; Coefficient -0.000; Std. Err. 0.001; t -0.740; P>|t| 0.462; 95% CI [-0.001, 0.001]
  - PRCL L3: Obs 7,011; R-squared 0.000; Coefficient 0.000; Std. Err. 0.001; t 0.540; P>|t| 0.588; 95% CI [-0.001, 0.001]
- Note: each regression includes one lag of real GDP growth at a time.

### Figures and diagnostic tests (Figures A1–A6)
- Figure A1: Symmetry in unconditional local projections — impulse responses of positive and negative shocks to state capacity and institutional quality on real GDP per capita growth; Year 1 refers to first year after shock at Year 0. Confidence intervals: 90% (darker) and 95% (lighter) based on small-sample corrected (SCC) standard errors. Panels disaggregate Advanced Economies and Emerging Markets and Developing Economies; left panel state capacity; right panels GEE and PRCL.
- Figure A2: Channels — state capacity shocks on employment (%) and labor force participation rate (pp.). Solid black lines show impulse responses; dark grey 90% SCC error bands; light grey 95% SCC error bands. Panels for Advanced Economies and Emerging Markets and Developing Economies.
- Figure A3: Channels — state capacity shocks on Inflation (%) and Overall Balance (pp. GDP). Same plotting conventions as Figure A2.
- Figure A4: Sensitivity — state capacity and institutional shocks on real GDP per capita growth under alternate specifications (Panel 1: country specific time trends; Panel 2: GDP growth expectations; Panel 3: Bias). Solid black lines plot impulse responses; light pink 90% SCC error bands. Solid blue line denotes unconditional baseline from Figure 5 with 90% SCC dotted blue bands.
- Figure A5: Alternative shock definitions — impulse responses of shocks on real GDP per capita growth; dark grey 90% SCC error bands; light grey 95% SCC error bands.
- Figure A6: Pre-trends test — reverse local projection assessing whether future GDP predicts contemporaneous shocks. The first difference of the MP shock is regressed on future real GDP growth (1 to 5 years ahead), controlling for lagged GDP growth, lagged shocks, lagged public debt-to-GDP, lagged consumer price inflation, and year fixed effects. Solid lines are coefficients on future GDP growth; shaded 90 percent confidence intervals with standard errors clustered at country level.

*Source: wpiea2025117-print-pdf - 19.087 IMF*

### References

### References (wpiea2025117-print-pdf)

### Foundational works on state capacity, institutions, and political order
- Acemoglu, D. (2005). Politics and economics in weak and strong states. Journal of Monetary Economics, 52(7), 1199–1226.
- Acemoglu, D., and Robinson, J. A. (2009). Foundations of state capacity. American Economic Review, 99(4), 1218–1244.
- Acemoglu, D., and Robinson, J. A. (2012). Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Crown Business.
- Acemoglu, D., and Robinson, J. A. (2019). The Narrow Corridor: States, Societies, and the Fate of Liberty. Penguin Press.
- Besley, T., and Persson, T. (2010). State capacity, conflict, and development. Econometrica, 78(1), 1–34.
- Besley, T. and Persson, T. (2009). The Origins of State Capacity: Property Rights, Taxation, and Politics. American Economic Review, 99(4), 1218-1244.
- Besley, T. and Persson, T. (2011). Fragile states and development policy. Journal of the European Economic Association, 9(3), 371-398.
- Besley, T. and Persson, T. (2014). The Causes and Consequences of Development Clusters: State Capacity, Peace, and Income. Annual Review of Economics, 6(1), 927- 949
- Hanson, J. K., and Sigman, R. (2021). Leviathan’s latent dimensions: Measuring state capacity for comparative political research. The Journal of Politics, 83(4), 1495–1510.
- O’Reilly, C., and Murphy, S. (2022). Dimensions of state capacity: A principal component approach to comparative governance. Comparative Political Studies, 55(7), 985–1011.
- Savoia, A. and Sen, K. (2015). Measurement, evolution, determinants, and consequences of state capacity: a review of recent research. Journal of Economic Surveys, 29(3), 441-458.
- Dincecco, M., and Katz, G. (2016). State capacity and long-run economic performance. Economic Journal, 126, 189–218.
- Dincecco, M. and Wang, Y. (2024). State Capacity in historical political economy. In Jenkins, J. A. and Rubin, J. (eds.) The Oxford Handbook of Historical Political Economy, Oxford University Press, Oxford.
- Heldring, L., and Robinson, J. A. (2017). Problematizing state capacity: The Rwandan case. African Affairs, 116(465), 717–743.
- Berwick, E. and Christia, F. (2018). State capacity redux: integrating classical and experimental contributions to an enduring debate. Annual Review of Political Science, 21, 71-91.

### Institutions, development, and economic growth
- Acemoglu, D., Johnson, S., and Robinson, J. A. (2001). The colonial origins of comparative development: An empirical investigation. American Economic Review, 91(5), 1369–1401.
- Acemoglu, D., Johnson, S., Robinson, J. and Thaicharoen, Y. (2003). Institutional causes, macroeconomic symptoms: volatility, crises and growth. Journal of Monetary Economics, 50(1), 49-123.
- Barro, R. J. (1991). Economic growth in a cross-section of countries. Quarterly Journal of Economics, 106(2), 407–443.
- Barro, R. J. (1995). Inflation and Economic Growth. Bank of England Quarterly Bulletin, 35(2), 166–176.
- Barro, R. J., and Becker, G. S. (1989). Fertility choice in a model of economic growth. Econometrica, 57(2), 481–501.
- Knack, S., and Keefer, P. (1995). Institutions and economic performance: Cross-country tests using alternative institutional measures. Economics and Politics, 7(3), 207–227.
- La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. W. (1999). The Quality of Government. Journal of Law, Economics, and Organization, 15(1), 222–279.
- North, D. C. (1990). Institutions, Institutional Change, and Economic Performance. Cambridge University Press.
- North, D, and R. Thomas (1973), “The Rise of the Western World: A New Economic History” (Cambridge: Cambridge University Press)
- North, D. (1991), “Institutions” Journal of Economic Perspectives, 5, 97-112.
- North, D. C., and Weingast, B. R. (1989). Constitutions and commitment: The evolution of institutions governing public choice in seventeenth-century England. Journal of Economic History, 49(4), 803–832.
- Rodrik, D., Subramanian, A., and Trebbi, F. (2004). Institutions rule: The primacy of institutions over geography and integration in economic development. Journal of Economic Growth, 9(2), 131–165.
- Greif, A. (2006). Institutions and the Path to the Modern Economy: Lessons from Medieval Trade. Cambridge University Press.

### Fiscal policy, multipliers, and public finance
- Alesina, A., and Perotti, R. (1996). Fiscal Discipline and the Budget Process. American Economic Review, 86(2), 401–407.
- Auerbach, A. J., and Gorodnichenko, Y. (2013). Fiscal Multipliers in Recession and Expansion. Fiscal Policy after the Financial Crisis, edited by A. Alesina and F. Giavazzi, University of Chicago Press, 63–98.
- Canova, F. (2024). Should We Trust Cross-Sectional Multiplier Estimates? Journal of Applied Econometrics, 39(4), 589-606.
- Miyamoto, W., Nguyen, T. L., and Sergeyev, D. (2018). Government Spending Multipliers under the Zero Lower Bound: Evidence from Japan. American Economic Journal: Macroeconomics, 10(3), 247–277.
- Ramey, V. A., and Zubairy, S. (2018). Government spending multipliers in good times and in bad: Evidence from US historical data. Journal of Political Economy, 126(2), 850–901.
- Romer, C. D., and Romer, D. H. (2019). Fiscal Space and the Aftermath of Financial Crises: How It Matters and Why. Brookings Papers on Economic Activity, 2019(1), 239–331.
- Tanzi, V., and Davoodi, H. (1997). Corruption, Public Investment, and Growth. IMF Working Paper, WP/97/139. International Monetary Fund.
- Tanzi, V., and Schuknecht, L. (2000). Public Spending in the 20th Century: A Global Perspective. Cambridge University Press.

### Labor markets, reforms, and macroeconomic interactions
- Bassanini, A., and Duval, R. (2009). Unemployment, institutions, and reform complementarities: Reassessing the aggregate evidence for OECD countries. Oxford Review of Economic Policy, 25(1), 40–59.
- Campolmi, A., and Faia, E. (2011). Labor market institutions and inflation volatility in the euro area. Journal of Economic Dynamics and Control, 35(5), 793–812.
- Gnocchi, S., Lagerborg, A., and Pappa, E. (2015). Do labor market institutions matter for business cycles? Journal of Economic Dynamics and Control, 51, 299–317.
- Parlevliet, J., and Kremers, J. J. M. (2018). Structural reforms and labor market outcomes: A review of the evidence and implications for the euro area. IMF Working Paper No. 18/45.
- Thomas, C., and Zanetti, F. (2009). Labor market reform and price stability: An application to the euro area. Journal of Monetary Economics, 56(6), 885–899.
- Zanetti, F. (2011). Labour market institutions and aggregate fluctuations in a search and matching model. European Economic Review, 55(5), 644–658.
- de Haan, J., and Wiese, R. (2022). Structural reforms: A review of the literature and new evidence. European Journal of Political Economy, 72, 102079.
- de Haan, J., Wiese, R. (2022). The Impact of Product and Labour Market Reform on Growth: Evidence for OECD Countries Based on Local Projections. Journal of Applied Econometrics, 37, 746–770.

### Methods, econometrics, and identification (local projections, panels, impulse responses)
- Jordà, Ò. (2005). Estimation and inference of impulse responses by local projections. American Economic Review, 95(1), 161–182.
- Jordà, Ò., and Taylor, A. M. (2016). The time for austerity: Estimating the average treatment effect of fiscal policy. Economic Journal, 126(590), 219–255.
- Plagborg-Møller, M., and Wolf, C. K. (2021). Local Projections and VARs Estimate the Same Impulse Responses. Econometrica, 89(2), 955–980.
- Teulings, C. N., and Zubanov, N. (2014). Is economic recovery a myth? Robust estimation of impulse responses. Journal of Applied Econometrics, 29(3), 497–514.
- Gonçalves, S., Herrera, A. M., Kilian, L., Pesavento, E. (2024). “State-dependent Local Projections.” Journal of Econometrics, 244(2), 105702.
- Herbst, E., and Johannesen, N. (2024). Revisiting Bias in Local Projections: Implications for Macroeconomic Policy. Journal of Economic Dynamics and Control, forthcoming.
- Driscoll, J. C., and Kraay, A. C. (1998). Consistent Covariance Matrix Estimation with Spatially Dependent Panel Data. The Review of Economics and Statistics, 80(4), 549–560.
- Pesaran, M. H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4), 967–1012.
- Pesaran, M. H. (2015). Testing weak cross-sectional dependence in large panels. Econometric Reviews, 34(6–10), 1089–1117.
- Granger, C. W. J., and Teräsvirta, T. (1993). Modelling Nonlinear Economic Relationships. Oxford University Press.

### Political economy, conflict, repression, and state failure
- Collier, P. (2007). The Bottom Billion: Why the Poorest Countries Are Failing and What Can Be Done About It. Oxford University Press.
- Collier, P., and Hoeffler, A. (2004). Greed and grievance in civil war. Oxford Economic Papers, 56(4), 563–595.
- Davenport, C. (2007). State Repression and Political Order. Annual Review of Political Science, 10, 1-23.
- Bates, R. H. (2008). When Things Fell Apart: State Failure in Late-Century Africa. Cambridge University Press.
- Fukuyama, F. (2004). The imperative of state-building. Journal of Democracy, 15(2), 17–31.
- Fukuyama, F. (2011). The Origins of Political Order: From Prehuman Times to the French Revolution. Farrar, Straus and Giroux.
- Fukuyama, F. (2014). Political Order and Political Decay: From the Industrial Revolution to the Globalization of Democracy. Farrar, Straus and Giroux.
- Herbst, J. (2000). States and Power in Africa: Comparative Lessons in Authority and Control. Princeton University Press.
- Migdal, J. (1988). Strong societies and weak states : state-society relations and state capabilities in the Third World. Princeton University Press, Princeton.
- Tilly, C. (1990). Coercion, Capital, and European States, AD 990–1992. Blackwell.
- Ross, M. L. (2001). Does Oil Hinder Democracy? World Politics, 53(3), 325–361.

### Industrial policy, development strategies, and historical perspectives
- Amsden, A. (1989), Asia’s Next Giant: South Korea and Late Industrialization, Oxford University Press.
- Evans, P. (1995). Embedded Autonomy: States and Industrial Transformation. Princeton University Press.
- Evans, P., and Rauch, J. E. (1999). Bureaucracy and growth: A cross-national analysis of the effects of "Weberian" state structures on economic growth. American Sociological Review, 64(5), 748–765.
- Johnson, C. (1982). MITI and the Japanese Miracle. Stanford University Press.
- Johnson, N. D. and Koyama, M. (2017). States and economic growth: capacity and constraints. Explorations in Economic History, 64, 1-20.
- Rodrik, D. (2004). Industrial policy for the twenty-first century. CEPR Discussion Paper No. 4767. Centre for Economic Policy Research.
- Wade, R. H. (1990). Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton University Press.
- Chang, H.-J. (2003). Globalization, Economic Development, and the Role of the State. Zed Books.
- Ogilvie, S. (2022). State capacity and economic growth: Cautionary tales from history. National Institute Economic Review, 262, R28–R50.
- Greif, A. (2006). Institutions and the Path to the Modern Economy: Lessons from Medieval Trade. Cambridge University Press.

### Trade, demography, and macroeconomic determinants of growth
- Frankel, J. A., and Romer, D. (1999). Does Trade Cause Growth? American Economic Review, 89(3), 379–399.
- Sachs, J. D., and Warner, A. M. (1995). Economic Reform and the Process of Global Integration. Brookings Papers on Economic Activity, 1995(1), 1–118.
- Bloom, D. E., and Williamson, J. G. (1998). Demographic transitions and economic miracles in emerging Asia. World Bank Economic Review, 12(3), 419–455.
- Fischer, S. (1993). The role of macroeconomic factors in growth. Journal of Monetary Economics, 32(3), 485–512.
- Solow, R. M. (1956). A Contribution to the Theory of Economic Growth. Quarterly Journal of Economics, 70(1), 65–94.
- Temple, J. (1999), “The New Growth Evidence,” Journal of Economic Literature, 37, 112–56.

### Corruption, regulation, and governance
- Mauro, P. (1995). Corruption and growth. Quarterly Journal of Economics, 110(3), 681–712.
- Djankov, S., La Porta, R., Lopez-de-Silanes, F., and Shleifer, A. (2002). The regulation of entry. Quarterly Journal of Economics, 117(1), 1–37.
- La Porta, R., Lopez-de-Silanes, F., Shleifer, A., and Vishny, R. W. (1999). The Quality of Government. Journal of Law, Economics, and Organization, 15(1), 222–279.

### Recent contributions on sovereign risk, indebtedness, and COVID-era effects
- Ortmans, A., Tripier F., (2021). COVID-induced Sovereign Risk in the Euro Area: When Did the ECB Stop the Spread? European Economic Review, 137, 103809.
- Choi, S., and Shin, Y. (2023). Household Indebtedness and the Macroeconomic Effects of Tax Changes. Journal of Economic Behavior and Organization, 209, 22-52.
- Romer, C. D., and Romer, D. H. (2019). Fiscal Space and the Aftermath of Financial Crises: How It Matters and Why. Brookings Papers on Economic Activity, 2019(1), 239–331.

*International Monetary Fund — References section from wpiea2025117-print-pdf*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025117-print-pdf.pdf_